|
Radiation therapy and surgical treatment and chemotherapy , consisting of the tumor three major treatment . IMRT (intensity-modulated radiotherapy, IMRT) appear , and the further development of the classic three-dimensional conformal radiation therapy techniques , radiation oncology in the history of the major changes . But in view of the complexity of the practical problems , IMRT treatment plans in the production, there are still many problems to be solved . This article closely integrated clinical needs , in order to achieve and improve conventional IMRT optimization techniques based around the the IMRT beam direction optimization (beam angle optimization, BAO) conducted in-depth research . The main work is as follows : ( 1 ) fully integrated particle swarm algorithm (particle swarm optimization, PSO) and genetic algorithms ( genetic algorithm , GA ) algorithm has the advantage , for IMRT optimization beam direction , proposed to approximate ( approaching) is a major evolutionary thinking of approaching genetic algorithm (approaching genetic algorithm, AGA) to automatically select the beam angle . Designed as the main means to approximate the search for ways and unique update policies to improve the search performance of the algorithm . The research results show that the simulated case and the actual clinical cases using AGA is feasible and effective , and compared with the GA more quickly find the optimal treatment plan to optimize the selection of beam angles . ( 2 ) combined with the knowledge of the evolution of biological DNA , a novel DNA genetic algorithm (DNA genetic algorithm, DNA-GA) to solve the BAO problem . Combined with the phenomenon of evolution of DNA variation hot spots (hot spots) and cold spots (cold spots) , the effective use of encoding location information , creative design code bit mutation operation ; same , DNA evolutionary process delete , insert , and inversion carefully designed to be the structure of the DNA-GA , delete, insert , and inversion mutation operation . DNA-GA mutation operation changes the the mutation genetic manipulation in the GA inside the auxiliary role to enhance the search capabilities of the algorithm . The preliminary results of the study showed that the DNA-GA is feasible and effective to solve the BAO problem and faster convergence than GA .
|